200-301 — frequently asked questions

Input errors is an aggregate counter that includes CRC, frame, overrun, and ignore errors, so any CRC error automatically raises it too.

Runts and giants count frames that violate the 64/1518-byte size limits; an FCS failure can occur on a frame of normal size.

In 10.2.1.0/30 the addresses are network .0, hosts .1 and .2, and broadcast .3, so .3 cannot be assigned to an interface.

No, both use 255.255.255.248, which is a /29 mask giving eight addresses, not the /30 required.

No Cisco PoE feature disconnects a device for under-drawing power; unused allocation simply remains available. PoE monitoring only reacts when a port exceeds its administrative power maximum, triggering syslog and err-disabled.

Option A describes PoE monitoring and policing, not classification override itself, which makes the switch ignore the PD's reported class and allocate a fixed amount. It is still the only option stating a true PoE behavior, so it is the intended answer.

EUI-64 is the very process that builds the interface ID from the MAC address. Disabling it removes the MAC-derived ID, so no MAC-based global unicast address would be created.

No. Stateful DHCPv6 assigns addresses from a server pool, and the interface ID does not need to match the MAC address, so it does not satisfy the question's requirement.

Because all 8 floors share ONE subnet, so the subnet must hold up to 320 addresses at once. A /25 gives only 126 usable hosts, so it would exhaust after roughly three floors.

Both would connect the office, but 255.255.0.0 (/16) reserves 65,534 usable hosts when only about 320 are needed. The /23 wastes nothing unnecessary, which is what "use address space efficiently" asks for.

On a real Cisco device you would assign the first usable address, 192.168.0.1/23. Every option here lists the network ID, so the intended discriminator is the subnet mask, and only B has the correct one.

Connecting servers and clients to a network is done by an intermediate device such as a switch; the endpoint is the server or client itself.

Forwarding traffic between VLANs requires a Layer 3 switch or router — devices that carry traffic between endpoints, not endpoints themselves.

The merge is between the core and distribution layers; the access layer remains a separate tier connecting end devices to the combined layer.

No. Its functions are not removed — they are performed by the combined core/distribution devices, so routing policy and distribution duties still occur.

Auto grants power only if budget remains, on a first-come, first-served basis. Because nothing is reserved, a detected device can be denied power, so auto fails the 'guarantees' requirement.

Static mode pre-allocates the configured wattage for the port even with no device attached, so any powered device drawing at or below that maximum is guaranteed to power up on connection.

Routing protocols and FIB construction run in the control plane; the data plane only consumes that prebuilt information to move packets onward.

Yes, hardware forwarding does consult the FIB, but option C names a sub-step while B states the data plane's overall responsibility, which is the intended answer.

Because the wording ties the behavior to a reload, implying the command does nothing until the switch restarts. PortFast takes effect immediately after configuration and on every boot, so the qualifier makes option B incorrect.

No. PortFast changes only the state-transition timing; the port still sends and receives BPDUs, and receiving a BPDU on a PortFast edge port causes it to fall back to normal STP behavior.

No. PortFast belongs only on access ports facing end stations such as PCs, phones, or servers; applying it to a trunk toward another switch risks creating a Layer 2 loop.

Packet Tracer simplifies CDP output, but Cisco's configuration guide says another interface's IP is substituted. The exam expects the documented behavior, not the simulator's display.

No. CDP is a Layer 2 protocol that exchanges multicast frames without any IP addressing, so neighbor discovery continues normally regardless of interface IPs.

The configuration revision is a per-switch counter, so whichever switch holds the highest revision wins the exchange regardless of its role. That is why a client with an inflated revision can overwrite a server's VLAN database.

No. VTP advertisements flow over trunk links that can be statically configured with switchport mode trunk; Cisco even notes trunk negotiation does not work across different VTP domains.

No. MAC learning records MAC-to-port mappings in the Layer 2 MAC (CAM) table; the ARP table maps IP addresses to MAC addresses and is built by ARP exchanges on hosts and routers, not by switch MAC learning.

By default the switch floods the frame out all other active ports in the same VLAN (unknown unicast flooding) to locate the destination. It is only dropped if a feature such as port security or an unknown-unicast drop is configured.

Because every other option is factually false: learning is enabled by default, unknown unicast is flooded not dropped, and MAC learning never populates the ARP table. D is the only true statement about the MAC address table.

Disassociation Imminent is a controller-initiated nudge inside a BSS Transition Management Request that gives a client a countdown before it is disassociated. The client still has to find and scan for a new AP, so roam latency is not reduced.

Only clients that support 802.11k use the neighbor list; older or non-compliant devices still scan all channels. In practice the feature benefits the vast majority of modern mobile devices, which is why it fits the all-device requirement.

Stateful (SSO) failover is a WLC high-availability feature between primary and standby controllers, preserving session state. LAG only bundles ports on a single controller's uplink and provides link redundancy, not session failover.

No. All member ports of a WLC LAG must share the same speed, duplex, and VLAN configuration and be connected to the same switch, since the bundle acts as one logical interface.

It is a common use, but it is narrower than the RFC 7519 definition. JWTs carry any claims — roles, expiry, issuer — so 'securely exchange information' is the accurate general description.

No. A standard JWT payload is Base64URL-encoded and digitally signed (JWS), so it can be read by anyone holding it. Encryption only occurs if the JWE structure is deliberately used.

Cisco IOS does not provide an 'lldp enable' command. Instead, LLDP on an interface is turned on using two separate commands: 'lldp transmit' to send LLDP advertisements and 'lldp receive' to accept them.

CDP is Cisco-proprietary and won't interoperate with non-Cisco VoIP phones or switches. Because the scenario is explicitly multivendor, disabling CDP globally and relying on the open-standard LLDP is the appropriate choice.

An EAPOL-Key frame is a single message, not a protocol stage. Those frames are the messages exchanged inside the four-way handshake, so the handshake itself is the step being named.

PMKID caching is a fast-resumption feature that reuses a cached master key. Since the question disables session resumption, a full EAP-TLS exchange occurs and no PMKID is stored or matched.

192.168.10.2 is in the clients' own subnet, so no relay is needed or possible — a same-subnet DHCP server is reached by broadcast. The helper address must be the remote DHCP server, 192.168.20.2.

192.168.20.1 is the gateway address of the server subnet, not the DHCP server itself. ip helper-address must point to the unicast address of the device that actually answers DHCP requests.

Catalyst switches can hash Layer 3 fields in hardware, and Cisco's WLC best-practices note recommends source-destination IP for more even distribution across many clients.

Hashing on source-destination IP keeps each client-to-controller flow on one link, preserving frame order while balancing aggregate traffic across the bundle.

Centralizing management tasks implies a controller-based or SDN architecture. A traditionally managed network is administered device-by-device, so centralized automation would change its management model rather than reduce costs within it.

Adding staff increases labor costs directly, which contradicts the goal of reducing maintenance costs. Cost-reduction answers on this exam virtually always involve automation or tooling, not more headcount.

Active enables LACP unconditionally by sending packets immediately. Passive waits until it detects LACP packets from the other side before enabling.

Auto is used for PAgP (Cisco proprietary), while Passive is used for LACP (IEEE standard). They operate on different protocols.

CDP is a Cisco proprietary protocol; it cannot communicate with or discover neighbors running non-Cisco operating systems.

No, unlike CDP which is typically enabled by default, LLDP must be manually enabled globally using the 'lldp run' command.

Authentication verifies who the user is (identity), while Authorization determines what resources the user is allowed to access (permissions).

Accounting logs all user activities and resource usage for billing, auditing, and security monitoring purposes.

No. Flooding is scoped to the VLAN where the frame belongs; ports in other VLANs are separate broadcast domains and do not receive that flooded frame.

Option D omits the ingress-port exception. A switch floods to every other port in the VLAN, not back out the port where the frame was received.

No. Flooding is the switch forwarding behavior; it can be triggered by unknown unicast, broadcast, or multicast frames. Broadcast is one destination type that causes flooding.

A /25 divides the last octet at 128: 192.168.25.100 is in 192.168.25.0/25, while 192.168.25.128 is the network address of 192.168.25.128/25, not a usable host in that subnet.

With /24, the range is 192.168.25.0–192.168.25.255; 192.168.25.100 and 192.168.25.128 are both host addresses inside it, so they use the same Layer 3 subnet.

Under /27, /29 or /30, 192.168.25.128 is a network address for its block, and .100 falls in a different block; only /24 creates a shared subnet containing both.

No. CDP typically exposes VLAN-related details such as the native VLAN and VTP domain, not the full VLAN database; option C is still the exam's accepted VLAN-information use.

CDP detail may include an entry or management address, but the two intended uses here are native-VLAN discovery and Layer 2 verification; Telnet access is not a core CDP reason and is insecure.

CDP frames are Layer 2, so a directly connected Cisco neighbor can still appear when IP addressing or routing is down; seeing the neighbor confirms the link is passing frames.

No. Cisco IOS devices require the global lldp run command because LLDP is disabled by default, unlike CDP, which is normally enabled automatically.

Yes. The default advertisement interval is 30 seconds, but the global lldp timer command changes it, and lldp holdtime changes the multiplier that determines neighbor hold time.

30 seconds is only the default. Because lldp timer can be set to another value, the frequency is configurable rather than fixed.

Yes. Repeating the same PUT request with the same body leaves the target resource in the same state, whereas POST is the usual non-idempotent create or submit verb. B contradicts PUT's core property.

Strictly, PUT updates an HTTP resource, not DNS zone data directly. A is the least-wrong choice because it describes an update, while B says non-idempotent and C and D describe read-only GET behavior.

GET is the read-only method browsers use to retrieve pages and display websites, so C and D do not describe PUT.

The use of a single wavelength of light, together with long-distance, low-loss transmission. Those traits belong to single-mode fiber, not copper.

Copper cabling uses a metallic conductor; bedding and sheathing are part of its physical construction. Single-mode fiber uses a glass core and cladding, so a conductor is the wrong match.

No. That trait describes the common short-reach copper deployment in this item; single-mode fiber is associated with longer distances and lower loss.

LACP (Link Aggregation Control Protocol) is an IEEE open standard (802.3ad) supported by most vendors. PAgP (Port Aggregation Protocol) is a Cisco proprietary protocol. The question specifies 'open-standard,' so LACP is the only valid choice.

By default, Port-Channel interfaces operate as Layer 2 switch ports. The 'no switchport' command converts the interface into a routed port, allowing you to assign an IP address and perform Layer 3 forwarding.

Yes, 'mode on' creates a static EtherChannel without negotiation. However, it is not an 'open-standard' protocol because it doesn't use LACP or PAgP. It also requires both sides to be manually configured as 'on'.

While PoE is necessary for APs, deploying switches is often a pre-deployment infrastructure task. The question focuses on tasks associated with the WLC deployment itself, such as logical configuration.

Yes, wireless traffic needs to be mapped to VLANs on the wired network for segmentation and routing. This is a core function of the WLC.

Check the third octet. If it is even, the network ID ends in that number. If it is odd, subtract 1. For example, 10.70.148.1/23 has network 10.70.148.0.

While the mask /27 fits 24 hosts, the IP 10.70.147.17 is not the first usable address. The network is 10.70.147.0, so the first usable IP is 10.70.147.1.